Any treatment strategy for HIV disease must confront a massive viral infection which is widely disseminated in blood, secondary lymphoid organs and other body compartments. It is now becoming apparent that these reservoirs exhibit different features in terms of viral burden and kinetics of virus and infected cell turnover. Recent estimates suggest that steady state levels of virus production are in the order of 10(9) viral particles per day with a plasma half life of two days. Correspondingly the production/destruction of HlV susceptible CD4 cells on average is also in the order of 10(9) cells per day. This places enormous stress on the immune system which must possess a remarkable regeneratiVe capacity in the face of the viral onslaught on the CD4 cell pool and other immune elements. Certain antiretroviral agents are able to reduce the viral burden by 99 to 99.9 percent and this decline is accompanied by a corresponding increase in CD4 cells. However, both effects are transient and reversed by drug resistant mutants which rapidly emerge and re-establish the steady state prior to treatment. There is considerable promise that combinations of anti-viral agents may significantly delay the appearance of drug resistant variants but one would still be faced with two major obstacles 1) the continuing presence of viral reservoirs that can spawn clinically relevant HlV variants able to repopulate the host and 2) the extent to which immune function in AIDS patients must be restored even if the virus can be controlled pharmacologically. The goals of this SPlRAT application are to reduce viral burden using combination inductive and maintenance anti-viral therapy. We will supplement the antiretrovirus therapy with alternative interventional strategies designed to reduce or eliminate residual viral reservoirs. Our major aim is then to reconstitute immune function through approximately HLA matched allografts of hematopoietic stem cells and thymus. This SPIRAT Program will include three Projects, two Scientific Cores and one Administrative Core.